From Phonology to Discourse: Studies in Six Colombian Languages

Coreguaje: Tone, Stress, and
Intonation

Frances L. Gralow

0. Introduction

Coreguaje' has a complicated suprasegmental system which, for a
number of years, caused much confusion to linguists studying the
language, who vacillated between calling it a multiple stress system and
a tonal system. (Of the other languages in the Western Tucanoan
family, some have been analyzed as stress languages and others as tone
languages.)

A major reason for the difficulty is the fact that pitch differences, as
well as stress differences, are very slight and very difficult to hear.
This is especially true in men's speech. Because mostly men were
available as language consultants, the first problem was hearing the
language well enough to record it correctly.

According to Eunice V. Pike, *' When high pitch is a feature of a tone
system, the syllable with high tone is frequently shorter than a syllable
with low tone.. .in analogous environments'' (1974:169). Since it is true
that high, stressed syllables in Coreguaje tend to be shorter in duration
than low, stressed syllables, this argued for a tone system. Eunice Pike
also stated, ''When high pitch is a contrastive feature of stress, vowel
quality is frequently conditioned by distribution into that stressed
syllable versus distribution into a nonhigh syllable. On the other hand,
when high pitch is a feature of a tone system, vowel quality is
unaffected by distribution into a high versus a nonhigh syllable'' (1974:
171). There is no conditioning in the quality of Coreguaje phonemes
which would correlate with the occurrence of high pitch on a syllable.
This was further evidence for a tone system.

Coreguaje nouns, the majority of which are disyllabic with the
pattern CVCV, present a problem because of the fact that in isolation
generally there is no contrast in either stress or pitch. In the end we
found that in certain frames there were four contrasting sets, but in
isolation phrase stress completely neutralized the contrasts, at least in
CVCV nouns.

4 Gralow

In the process of the analysis it finally became evident that stress
was basically predictable, though tone was not. When no consistent
tone sandhi patterns were discovered, the function of intonation was
considered. It was then found that Coreguaje has contrastive intonation
patterns which affect and even override word tone patterns, making the
latter very difficult to distinguish.

1. Stress and Rhythm

Rhythm is an important consideration in Coreguaje phonology. Word
length, stress, and intonation all work together to affect Coreguaje
rhythm. (Intonation is discussed in section 3.)

Every phonological word, sentence-level phrase, and sentence in
Coreguaje has a tendency to fade away; that is, as a unit progresses,
the syllables tend to become lower in pitch, shorter in length, and
diminished in volume. The first syllable of a stem of a word is usually
stressed. Stress is characterized by perceived loudness and length.
High pitches tend to be higher toward the beginning of the word,
phrase, or sentence than they are toward the end. (For a slight
deviation from this pattern, see sect. 3.1.) Also, the final syllable or
syllables of independent verbs tend to become voiceless before pauses.

The length of the word, phrase, or sentence is very important in
Coreguaje phonology. The decrescendo effect is most prominent in
longer units. Word length is also important in tone rules as well as in
morphophonemic rules,

2. Tone

There are two tone levels in Coreguaje, high and low. (High tone is
written ' and low tone is unmarked.) Words of two syllables contrast in
all four possible sequences, e.g., high-high, hdma 'deer'; high-low,
ndso 'monkeys'; low-high, rut^a 'flies'; and low-low, wekH 'tapirs'.
These contrasts are found in the following contexts: (1) phrase initially
on words which are not in the phrase nucleus, or (2) phrase medially
on words which occur either following words where there is no
perturbation across word boundaries or following words ending in low
tone and which are in the same close-knit sequence.^ (For further
discussion on phrases, see sect. 3.)

All single morpheme verb stems and the majority of noun stems
consist of two syllables. Syllables are of the following types:

First syllable: Vi, CVi, or CVi'?
Second syllable: V2, or CV2
'^ = glottal stop

Coreguaje

C = all consonants except "^

Vi = all oral and nasalized vowels

V2 - all oral vowels^

Examples:

*to drink'
'fire'
'manioc'
'possum'

Tone rules vary slightly between stems of different syllable patterns.
The three stem types are (C)VCV, (OV^CV, and (C)VV.
Examples:

jpwa
hua

'to enter'

'to give advice'

'to pierce'

Stems of the first two patterns function essentially alike, except that
syllables with glottal stop tend to be higher in pitch than high tone
syllables without it. To date, no stems of the pattern (C)VV have been
found with the tone pattern low-high. ((C)VV stems are considered to
be two syllables.)

2,1- Nouns. The four contrasting tone patterns mentioned are partially
or completely neutralized when two-syllable nouns are pronounced in
isolation, e.g., as a one-word statement in response to the question
"What are they?" — hama 'Deer/ — or as a one-word clarification
question in response to a statement or question such as "(Did you say)
deer?" Basic tone patterns were determined by listening to nouns in

such close-knit phrases as pa'^imey 'There are ' and short

sentences such as re haamijPi '1 see a '

Nouns with the pattern (C)VCV show neutralization of tone when
pronounced in isolation. (In the following examples, * indicates stress
on the following syllable, ' indicates high pitch, a indicates high falling
pitch, and ' indicates low pitch.)

Basic Form

Statement

Question

*hama 'deer'

*hama 'Deer.'

*hama 'Deer?'

^ndsd 'monkeys'

*nasd 'Monkeys.

' *ndsd 'Monkeys?'

*rut^a 'flies'

*rut^a 'Flies.'

W4t^a 'Flies?'

*w€kH 'tapirs'

^\fekH 'Tapirs.'

\\^€k^i 'Tapirs?'

6 Gralow

Those nouns with the pattern (C)VV, however, do contrast
somewhat in isolation as a one-word statement.

Basic Form

Statement

Question

Wie 'anteaters'

*mie 'Anteaters/

Wie 'Anteaters?'

^sii) 'pheasants'

'a7// Theasants/

KsiTi * Pheasants?'

^t^eci 'a bird'

't^^a ^A bird/

7V« ^A bird?^

2.2. Verbs. In this section the tone patterns of independent verbs in
isolation are presented/ Since Coreguaje is a verb-heavy language,
sentences consisting of only an independent verb are very common.
Since tone is not conditioned across pauses, there is a relatively
uncomplicated body of data for consideration.

An independent verb must have at least one high tone and one low
tone. It cannot have more than three high tones. When two high tones
occur on contiguous syllables, there is some free variation in the actual
pitches so that one or the other may be slightly higher, or they may be
the same. When three high tones occur on contiguous syllables, if the
first or the last high-tone syllable occurs contiguous to a low-tone
syllable, it will usually be lowered to a mid. A very few suffixes have
inherent high tone. Sequences of tones on the independent verb are
determined by the interrelationships of the tones of the stem, the tones
of the suffixes, and the total number of syllables in the suffix cluster.

The following rules apply to independent verbs. Examples are
provided after each rule. Note that stems of the tone pattern high-low
often act differently from the others. (Here high tone is marked 'and
low tone is unmarked.)

Rule 1. The first syllable of a suffix or suffix cluster bears high tone.
Examples:

Sill

go

-I- -mi

masc sing

Sill mi

'he is going'

jawe +

be hidden

intent

-I- -mi

masc sing

jaweha'^mi
'it will be hidden'

hore -f -si -h -na -h - '-^ + -me = horesina '^me

lie complete pi anim nom stat pi 'they lied'

kahe H- -si'^ + -kH

-H -a^ -\~ -mi

= kah es i ^/c ^ia '^mi

descend complete masc sg nom stat masc sg 'he descended'

asa -+■ -repa + -ma '^ -H -mi
hear truly neg masc sg

= asarepama ^mi

'he truly did not hear'

Coreguaje 7

Rule 2. The last syllable of the independent verb cannot bear high tone
except when the suffix is only one syllable long and the verb
components fulfill one of the following conditions.

a. The stem is of the (C)VV syllable pattern and the last syllable of
the stem is low. Examples:

qi + -mi = qimi

eat masc sing 'he is eating'

peo + -kH = peokH

be not neuter interr Ms there not any?'

k^oo + -mf = k^oomi

receive masc sing 'he receives'

b. The suffix has inherent high tone. Examples:

jo'^o -\- 'kH = jo'^okH

work masc sing interr Ms he working?'

mit^d -^ 'kH = mk^akH

come masc sing interr 'is he coming?'

asa
hear

+ -i

masc sg compi interr

asm

'did he hear?'

If neither of the above conditions is present, the penultimate syllable
of the independent verb with a suffix cluster of two syllables bears high
tone and the ultimate syllable bears low tone. Examples:

Rule 2

masi + -ma '^

+

-kH = *masima'^kH

masimd^kH

know neg

masc sg
interr

'doesn't he know?

jawe

+

-mi = '^jdwemi

j awe mi

be hidden

masc sg

'it is hidden'

kahe

+

-mi = kdhemi

kdhemi

descend

masc sg

'he descends'

Rule 3. When there are four or more syllables in the suffix cluster,
Rule 2, of course, does not apply. If the stem is of the pattern high-
low, a high tone occurs on the second syllable of the suffix cluster in

8 Gralow

addition to the high tone on the first syllable (from Rule 1). Suffix
clusters following all other stems have a high tone only on the first
syllable of the suffix cluster, when all the suffixes are inherently low.
The tones of the stem remain the same as its basic form in these cases,
though the second syllable of a low-low stem may vary to mid.
Examples:

jo'^o + -repasina'^me = "^o'^orepasina'^me *they truly worked'

work they truly did *jd'^orepasina'^me Rule I

jd'^orepasina'^me Rule 3

k^oo -f -si'^kNa'^mi = *k^dosi'^k'*ia'^mi *he received'

receive he-completive *k^dosi'^kHa '^mi Rule I

k^oosV^kHa^^mi Rule 3

asa + -si'^kHa'^mi = *asasi'^kHa'^mi 'he heard'

hear he-completive asast'^kHa '^mi Rule 1

Two-syllable suffixes which have inherent high tone tend to act
differently from suffixes without inherent high tone. They also seem to
have stress on the first syllable as do stems. Historically these high
tone suffixes were probably verb stems, but because of their
dependence on other verbs they now function as suffixes. These
suffixes are:

-k'^ai benefactive

-hai accompanying movement

-koso probability

When used in an independent verb these suffixes keep their inherent
tones, and the basic tones of the stem do not change.

The suffix -kua behaves irregularly. When following a high-high
stem, the stem tones become high-low and -kua bears low tone. All
other stems become low-high and -kua is high-low. Suffixes following
-kua follow Rules 1-3.

3. Intonation

Two types of intonation contours have been observed in Coreguaje
speech: those covering a phrase (or sentence portion or clause), after
which there is a tentative pause, and those preceding a final pause.

(1) In nonfinal contours, there is an optional premargin in which
basic word tones of words are maintained, though the pitch levels are
usually quite close together. The nucleus of this pause group (defined
as the syllable containing the loudest stress and longest vowel length)
occurs on the stressed syllable of the last word in the group, i.e., on

Coreguaje 9

the last stressed syllable in the pause group. The final pitch of this unit
is predictable in accord with the tones of the final word or words as
follows:

If the final word has the tones high-high or low-high, there is no
change in pitch. If it is high-low, the low tone is raised to just slightly
lower than the high syllable, so that the word is pronounced high-mid.
If the word preceding a high-high final word in this group ends in low
tone, the intonation generally follows the tone patterns fairly closely. If
the penultimate word ends in a syllable bearing basic high tone, that
high tone may remain high, or the pitch may rise gradually from the
beginning of the penultimate word to the high of the nucleus, thus
giving the last syllable of the penultimate word a mid pitch.

9cai t\yi_ 'p^i^t^ 'what is in the body'
body being, life

!jdi^ ^haaW) 'when the witch doctor looks'

witch doctor when looks

*fear in one's heart'

or

(2) In final contours there is a strong down-glide of pitch beginning
with the nucleus of the word in the phrase nucleus and continuing over
the rest of the syllables in the phonological phrase. The nucleus of this
phrase is usually on the nucleus of the independent verb, but may also
occur on question words (which are at the beginning of the phrase) or
on any word which receives emphasis. All high tones occurring after
the phrase nucleus in the final contour are considerably lowered.

'noo/m^^^-Jro/ 'the witch doctor reveals'

he shows witch doctor

In addition, there is an optional postmargin to this main final phrase
which may be a sort of tag comment in the form of a dependent clause
or a repetition of one of the items in the main clause, such as the
subject or object. All words in this final margin are perceptually very

10 Gralow

low, though at times a very sHght rise is heard on basic high tone
syllables. The final word of this postmargin has an even lov^er down-
glide.

^ukHna

witch doctors they-know banisterium those who drink
'witch doctors know (those who drink banisterium)'

4. Summary

Coreguaje phonology is characterized by a strong decrescendo effect
over phrases, sentences, and longer words. This tends to minimize
contrasts between high and low tone, depending on where in the pause
group words occur. Nonfinal intonation allows phrase-final high tones
to remain high, but pulls corresponding low tones to mid. Final
intonation causes high tones to be lowered, often to the point of being
perceived as low. Otherwise, noun and verb stems of two syllables
contrast in all four possible sequences of tone. Suffixes with inherent
high tone generally keep those tones. Suffixes with no inherent high
tones may receive one or more high tones depending on the tone
pattern of the stem and the total number of syllables in the word.

Notes

1 Coreguaje is a member of the Western Tucanoan language family and is spoken by
approximately 800 people who live on the tributaries of the Orteguaza River below
Florencia and the Caqueta River below Solano, in Caqueta, Colombia. Research for this
paper was done under the auspices of the Summer Institute of Linguistics between
February 1975 and August 1979. Most of the present analysis was developed during a
linguistic workshop conducted by Dr. Donald A. Burquest of the University of Texas at
Arlington in August 1978 at the Summer Institute of Linguistics center in Lomalinda,
Colombia. It was then revised during a phonology workshop conducted by Dr. Ruth
Brend of Michigan State University in February 1981, also at Lomalinda. The principal
Coreguaje language consultants in this project were Alicia Garcia Pizarro and Silvio
Fajardo from Maticuru and Elicenia Gutierrez Gasca from Jacome.

2 In close-knit verb phrases where the first word of the phrase is an adverb, a verb stem,
or an incorporated noun, the first syllable of the following conjugated verb is raised to
high tone if the first word ends in low-low.

Examples:

'q) + Svesim^ = 'q) 'wesimi 'he doesn't know how to eat'

eat he doesn't know

'ho'^e + 'ra)mi = 'ho'^e 'ra)m^ 'he's coming again'

again he comes

hva'^) + 'sa)mi = Sva'^} 'satmi 'he's going hunting'

meat he goes

Coreguaje 1 1

The only other perturbation occurring other than that conditioned by intonation is the
phenomenon of either a high tone or a low tone optionally lowering or raising to mid
when occurring between a low and a high.

3 If Vi is not the same as V2 (disregarding nasalization), they cannot both be rounded
vowels. To avoid this combination, u becomes f in the environment of o,

4 This section is included in outline form in the article by Young, Cook, and Gralow
about Coreguaje phonology.

References

Pike, Eunice V. 1974. ''A Multiple Stress System Versus a Tone

System." UAL 40:3:169-75.
Young, Carolyn M., Dorothy Cook, and Frances Gralow. "Informe

Descriptivo de los Fonemas del Coreguaje." Sistemas Fonologicos

de Idiomas Colombianos 5.

LANGUAGE DATA
Amerindian Series No. 9

From Phonology to Discourse:

Studies in Six Colombian Languages

Ruth M. Brend

Editor

Summer Institute of Linguistics

7500 W. Camp Wisdom Road

Dallas, Texas

1985

Editor-in-chief Desmond Derbyshire

Asian- Pacific Series Phyllis Healey
David Thomas
African Series Pamela Bendor-Samuel

Amerindian Series Viola Waterhouse

©1985 by the Summer Institute of Linguistics
ISBN 0-88312-091-7